Planning Strategy Representation in DoLittle

This paper introduces multi-strategy planning and describes its implementation in the DoLittle system, which can combine many di erent planning strategies, including means-ends analysis, macro-based planning, abstraction-based planning (reduced and relaxed), and casebased planning on a single problem. Planning strategies are de ned as methods to reduce the search space by exploiting some assumptions (socalled planning biases) about the problem domain. General operators are generalizations of standard Strips operators that conveniently represent many di erent planning strategies. The focus of this work is to develop a representation weak enough to represent a wide variety of different strategies, but still strong enough to emulate them. The search control method applies di erent general operators based on a strongest rst principle; planning biases that are expected to lead to small search spaces are tried rst. An empirical evaluation in three domains showed that multi-strategy planning performed signi cantly better than the best single strategy planners in these domains.